WAI Authoring Tool Guidelines Working Group
Date: 24 September
Time: 12 noon Boston time (what time is that where I am?)
Phone number: W3C MIT Bridge, +1 (617) 258 7910
we will be there, start of the week. It is possible we will withdraw. We need to see groups that will be there so we can make sure.
JT schedule OKed yet?
CMN Daniel is organising - w should know by end of this week
JT Vancouver is our other possibility - before or after.We preferred after plenary.
(CMN is halfway through making plenary minutes readable)
CMN Next week ER are in a face to face so no joint meeting they proposed joint telecon 15 oct.
JT Do we want to drop monthly meetings?
CMN this is a one-time replacement except that next month wendy and CMN will not be available for joint telecon scheduled.(i.e. november)
JT We should decide on November at next joint call.
Resolved: Decide on future joint calls with ERT group at next joint call, mid-October
LN hello
WL: jutta welcomes you as do I (William)
phone: WL == Will Loughborough, HS == Heather Swayne
KHS == Katie Haritos-Shea, CMN == chaals, JT == Jutta
WL email clients used for archived lists in particular
JT It is really the archiving functionality.
WL that dictates if the tool is an authoring tool or not
JT it isn't the client that has control over the archives.
WL It has control over formatting, linking, etc.
JT Does archive inherit everything from email message?
CMN I thin it is the archiving tool that is the one we need.
JR With Hotmail etc there are lots of people using email as web content.
CMN Yep, that's true.
WL We should agree whether this is desirable, and then is it feasible?
JR Any mail client can count.
JT Who controls the rendering?
JR Email clients allow you to write HTML docs, send as email, they have to make sure it is done right. User Agent guidelines apply to the bits in the middle to some extent, but also under authoring tools.
CMN I think we are looking at complex systems, so techniques will reflect that.
JT how do we keep scope down?
JR I think the scope is broad.
JT Do we want specific techniques?
JR I think a lot of it is covered already
CMN agree with Jan - bunch of this is covered, but there are probably some some user interface techniques that we need to think about.
JT it is also what is done to render ordinary mail - re-processing.
CMN i.e. conversion tools
JR Archiving tool - do we tell it not to accept inaccessible tools, send back a message asking for repair?
WL That might be too late.
CMN many lists have moderation - it could be semi-automated and the tool could refer back to author before publishing.
KHS would be easier to hav it on the mail client.
WL First question is whether this is in our scope?
JT Authoring tools include these things. Then we need to look at techniques.
CMN I think this is in our scope, we need to work on our techniques.
KHS e.g. an email program could help you make stuff better.
WL parts of this are under User Agent guidelines too.
JT We are thinking this is in our scope.
Resolved: Email clients and archiving systems are in our scope
JT So where do we go? What kind of techniques?
Action KHS: Do sample techniques for Outlook.
CMN sent a message to list outlining requirements we were trying to meet.
First question: Are there any others I didn't put in there?
proposed: We have started work on a new document to meet existing requirements, if further things need to be done we wil add them as we go.
JR Current proposal seems to cover where we are at.
JT What next?
CMN I need to publish as a document, we need to agree on it.
CMN We need permission to publish a first draft of a new spec. This will help us show why we needed one, and why we want to publish what we have.
Action CMN: Publish current list as requirements document.
WL Virtual conferencing comes under authoring tools.
CMN Agree. There are parts of this we are looking at already under courseware so it is about making sure we have each piece, and identify clearly how to deal with them
JT Big challenge is that they are synchronous
WL This is going to be very popular now to replace airplanes..
JT GJR and I are wokring in this area on IMS maybe we can share that work here
CMN Yes. Please
Action JT: Get permission to share the IMS work on courseware with this group, and then do so.
JT LN is also part of that group.
LN I am wondering if we should think about sharing all IMS stuff - for the benefit of IMS as well as W3C - the more editing the better, surely?
JT How do we deal with revisions of products, revisions of reviews etc? We have a fairly informal approach, on our reviews page Jan pointed out that some developers modify things without changing version numbers.
CMN there is a tension between giving developers time to read reviews, and having timely information output.
JT Right. We probably should have a disclaimer about versions being updated.
WL Problem is that there are long development cycles, but people can tweak them to claim they have changed a particular fault or say they have changed something in a new version.
JT I don't think we can accept "this is different in the new version" without seeing it.
CMN, JR right.
JT we have excluded comment in the past on that.
WL If there is a two-year lag there is two years when the inadequate version is available.
JT we should set timelines. How long do we allow developers to respond? what if they don't? Do we just publish? how long would we hold off to take account of new work?
CMN If there is no response, we go ahead and publish.
JT Do we want to contact..
CMN 4 weeks is enough for them to respond.
JT 2 responses. One is there is an error in the evaluation, other is "we will fix please review new version"
CMN If there is an error in evaluation, we should fix it before publishing, if there is a disagreement in interpreting the guidelines we should publish both opinions. If the developer asks to hold off for new version I think it should be only if the new version is avilable in a few weeks
JT 4 week limit?
LN and error is a legal problem so you should ensure they have had chance to check it -
CMN they can update the review themselves when they have produced a new version.
CMN SHould we start with 4 weeks from finish to publication?
WL theater and book reviewers don't do that
JR What if we did make a mistake?
LN to WL they are not making tech assessments
WL: lid: yes they are when they speak of "bad lighting" etc.
CMN If there is a mistake in a review we should try to correct it.
JR do we keep legacy reviews?
LN no, bad is opinion
CMN I think we should keep legacy reviews but point to most up-to-date.
LN you are claiming to be asserting facts
WL: so are tech assessments
JT Do we agree that if there are differences on interpretation should we publish both sides?
Resolved: If there are different interpretations of ATAG in a tool evaluation we will publish both of them.
WL: Nope, evaluations
JT For developers saying "next version" should we publish, wait, what?
JR can a company submit a review?
CMN anyone can.
CMN propose we wait four weeks for developers to do repairs, but then publish.
JR That's OK for online tools, but for software is it 4 weeks to alpha, 4 weeks to shelf, ...?
WL There is no stopping anyone else from publishing this how they like. We are furnishing a template.
JT We are putting it on our site, and if we are going to publish this...
JR Say someone does an eval, finds a problem, says they will have a beta in 4.5 weeks
CMN This is a recommendation.
JT Recommendations are 4 weeks to respond, plus four weeks for new version?
JR CMN sure
WL We want to understand the developers, but also user needs. Are we going to have a 4 week constraint on someone whose job got crunched because they couldn't use the tool?
CMN For particular issues people can use mailing lists (and should mail to developers)
WL Are we going to have a 4 week delay before putting things on our site?
JR Think so. This isn't just a complaint site. 4 weeks is time to make sure it is factually correct
JT opportunity for developers to see and respond is another form of encouragement. if we publish straight out there is no sense of being able to defend their work. I think there is a lot of advantage in that one month.
HS FP 2002 review took us longer than 4 weeks to get a response. haven't had follow up yet?
CMN I think 4 weeks is good
WL me too.
Resolved: we want to offer 4 weeks for review, and will wait 4 more weeks for new versions to be available where applicable.
Action someone: propose update for the text we have on our reviews page
Action CMN find Frontpage 2002 review and response.
WL Should people fill in a form?
JR We could put this into the template too.
WL Can we automate the evaluation process following the protocol?
JT not at the moment. We have been talking about a wizard to do conformance test.
WL Otherwise the information is going to disappear.
CMN There is a wizard tool for doing some of this, but I have been waiting for it to be checked a bit before pushing use of it. (And it does EARL)
Last Modified $Date: 2001/09/24 17:42:57 $